Decomposing a Chunk into Its Elements and Reorganizing Them As a New Chunk: The Two Different Sub-processes Underlying Insightful Chunk Decomposition

Xiaofei Wu, Mei He, Yinglu Zhou, Jing Xiao, Jing Luo

Familiar chunks can be processed highly efficiently, and this automatic process can prohibit the problem solver from developing novel and original ways to creatively solve difficult problems. For this reason, the role of the reverse process, chunk decomposition (CD), the process by which familiar patterns are broken down into their component elements in order to be regrouped in another meaningful manner, has been generally recognized as part of the creative process. However, previous studies on this issue have mainly focused on the decomposition process of CD (the D-process), while the reorganization process of CD has been greatly neglected or has not been distinctively identified in previous work. In this paper, we argue that the R-process could be equally as important as the D-process for CD. Even if a problem solver manages to decompose a familiar chunk into its elements, he or she still may not solve the problem if these elements are not successfully organized in a new and meaningful manner. To investigate whether the cognitive mechanism of the R-process is different from that of the D-process, we designed an experiment for detecting the effects of chunk tightness, which is regarded as the key factor in CD and which can be experimentally manipulated by the radical-level (loose) and stroke-level (tight) Chinese character CD tasks in the D-process, the R-process, and the more purified organization task (the O-process task) that does not involve the decomposition process. Our results showed that the stroke-level (tight) task was more difficult than the radical-level (loose) task for the D-process. However, for the R-process, the stroke- and radical-level tasks showed no differences in performance. Moreover, for the more purified reorganization task, the O-process task, the radical-level organization and reorganization could be even more difficult than the stroke-level organization and reorganization. This result demonstrated that the cognitive processes underlying chunk decomposition and reorganization are fundamentally different. Therefore, more general concepts such as chunk restructuring that could include both D- and R-processes might be more suitable in accounting for this type of creative insight.